OpenAI Launches Presence for Enterprise AI Agent Automation
OpenAI has introduced Presence, a new enterprise platform that helps organizations deploy AI agents across customer service and internal business operations. Instead of offering only foundation models, the company now provides a complete system that connects AI agents with enterprise data, workflows, and governance controls. The launch signals OpenAI’s broader push into enterprise software as businesses seek production-ready AI solutions.
Enterprise AI agents with built-in governance
Presence enables businesses to deploy AI agents for voice and chat interactions across customer support, sales, insurance claims, billing, and employee IT services. Moreover, organizations can define exactly what an agent may access, which actions it can perform, and when it must escalate a request to a human employee.
Rather than functioning as a standalone AI model, Presence combines company policies, standard operating procedures, approved actions, guardrails, simulations, and evaluation tools into a single deployment platform. As a result, enterprises can test agent behavior before launch and monitor performance after deployment.
OpenAI also integrates a Codex-powered improvement workflow that analyzes production sessions, identifies performance gaps, and proposes updates. However, organizations remain responsible for reviewing and approving every change before it reaches production.
Designed for production-scale deployments
According to OpenAI, each Presence deployment begins with a specific business workflow instead of a general-purpose assistant. Consequently, every AI agent receives access only to the systems, documents, and permissions required for its assigned role.
The company says the platform already powers its English-language phone support service. OpenAI reports that the deployment resolves approximately 75% of inbound support requests without human intervention. Furthermore, its continuous improvement process reduced human handoffs by 15 percentage points within ten days.
OpenAI also confirmed that companies including BBVA, SoftBank, and IAG are adopting the platform for enterprise customer service initiatives.
OpenAI moves beyond foundation models
With Presence, OpenAI is positioning itself as an enterprise software provider rather than only an AI model developer. Instead of requiring businesses to assemble multiple AI tools independently, the company delivers deployment services, governance frameworks, testing systems, and operational support through a single product.
The platform is currently available through a limited general availability program for eligible enterprise customers. OpenAI’s Forward Deployed Engineers and selected global systems integrators oversee deployments, while a self-service version has not yet been announced. Meanwhile, developers can continue building voice applications through the OpenAI API separately from Presence.
